THE ARGUMENT FROM EMBRYOLOGY 103 



acquired, or larval characters, and the features that 

 are really due to inheritance ; while the correct dis- 

 crimination of them is one of the greatest problems 

 which an embryologist has to solve. As a general 

 rule, secondary characters will be more variable, 

 because the different groups will probably have 

 acquired them independently. Again, secondarily 

 acquired characters must always be useful, must 

 confer some advantage on their possessors, or other- 

 wise they would never have been preserved ; while 

 on the other hand, structures such as rudimentary 

 organs, which are of no practical use, must be 

 inherited, for in no other way can their presence be 

 explained. 



One other cause of falsification of the ancestral 

 history in actual development may be briefly alluded 

 to. It happens not uncommonly that the larvae and 

 adults have entirely different habits, and in such 

 cases the transition from one to the other is not 

 always a gradual one, but may be effected by an 

 abrupt, almost violent metamorphosis. 



Take the case of a butterfly or moth. From the 

 egg emerges a caterpillar, a soft-bodied vermiform 

 animal with short fleshy legs adapted for crawling 

 along the branches and leaves of plants, with jaws 

 adapted for biting these leaves, and an alimentary 

 canal fitted to digest them as food. The caterpillar 

 feeds and grows rapidly, but retains its shape, and, 

 except in size, makes no appreciable approach 

 towards the adult condition. Having reached its 

 full size, it changes into a chrysalis or pupa, and during 

 this state, which lasts for weeks or months, it takes 



